The goal of our proposal is to identify proteins that serve as prognostic and treatment response biomarkers in childhood acute myeloid leukemia (AML) using specimens from patients enrolled in the Children's Oncology Group (COG) multicenter trials. Our ability to treat cancer more effectively depends critically on improving and refining tumor classifications. This unmet need is particularly acute in childhood AML. As part of the Therapeutically Applicable Research to Generate Effective Treatments (TARGET) Initiative, we are analyzing 150 pediatric AML samples with a battery of high throughput technologies in order to identify genomic alterations, epigenetic changes, mRNA and miRNA abundance signatures that may serve as prognostic biomarkers. Because critical aspects of a cell's physiology may be better captured by its protein complement, we propose to carry out proteomic analysis of the same patient cohort using a new method for correlating protein abundance across many samples that was developed in our laboratory. Most currently available proteome profiling technologies (e.g. 2-D gels, ICAT, SILAC) have limitations that prevent their use for profiling large numbers of specimens. We have developed a gel- and isotopic label- free platform for analysis of mass spectrometric data that is capable of profiling large numbers of clinical samples. Our preliminary results demonstrate that our method can identify and quantify a significant portion of the proteome across several hundred samples without resorting to fractionation. Furthermore, we have already demonstrated that this method can differentiate between acute myeloid and acute lymphoid leukemias using protein expression patterns. The proteome analysis of the TARGET cohort in this study will enable us to identify prognostic protein-derived signatures in childhood AML. Multi-analyte signatures will be obtained by integrating protein level measurements with other molecular signatures such as mutational, chromosomal and epigenetic changes obtained by comprehensive sample annotations derived through the TARGET program. Integration of proteomic and genomic signatures will improve our ability to stratify childhood AML sub-types and improve prognostic and therapeutic outcome assessment. PUBLIC HEALTH RELEVANCE: Acute myeloid leukemia is a heterogenous group of blood cancers that can be cured with chemotherapy alone or with chemotherapy followed by stem cell transplant. At the present time our ability to determine which patient will be cured with chemotherapy alone and which patient needs stem cell transplant is inadequate which may lead to application of toxic treatments in patients that do not need it or insufficient treatment for patients that need aggressive therapy. In this proposal we use a novel mass spectrometric technology to measure the abundance of approximately 1000 proteins across large number of diagnostic samples from children with acute leukemia in order to identify proteins and peptides that can predict patients' prognosis and thus assist physicians in determining the best treatment option for individual patients.